Thanks for the nice historical context of the mechanical governor. The discussion of different power sources, steam versus gas turbine vs hydroelectric penstock etc. and the reasons not to modulate power too quickly or too slowly was nicely done. This could have been a very dull and dry description, but you made it interesting. Overall, a governor is a feedback device. In electronics, we are always worried about oscillation with feedback devices, and at least one portion of the loop must have a substantially slower response than the prime mover to avoid oscillation. One can achieve oscillation if you meet the Nyquist criterion, and I will stop there as it is too mathematical to be interesting in a short comment.
Yikes! That is why the 100 KW 400 VDC generator we are designing and building for our two passenger 10 electric motor blown wing airplane uses electronic control of both the 143 HP turbonormalized engine and the active rectifier after the permanent magnet brushless generator. All those mechanical adjustments are too unreliable and too hard to modulate for flight.
Thank you, and yes oscillations are a problem, even in mechanical governors. It's one of the reasons you see droop in single units, the droop helps calm the mechanical action. As you take droop out the governor gets harder to tune. Most electronic governors are based on a PID loop, but it gets tricky because a governor swing can start swinging fuel pressure, and boost. Then you have all these other loops contributing to the problem. The PID triangle was there in mechanical and hydraulic governors too, but not as well defined.
For those readers not familiar with a PID controller, this a governor that uses the Position distance from a setpoint such as temperature, the Integral of the adjustment time from the deviation from the setpoint, and the Differential of rate of change returning to the setpoint. An example makes this more clear. I bought an industrial oven for a project that needed a precise 250° F to activate a single component epoxy to attach a metal disc to urethane. The urethane was rated to 250° F. The surplus oven did not work. I removed the old electronics and installed a tiny 2” by 2” learning PID temperature controller that had a digital display. The learning PID controller was industrial surplus and cost $100. I set it to 250° F. The first time was the learning run. The oven rapidly went to 251 degrees, then 249 degrees, then 250 degrees steady. The second time the oven was heated up the temperature went rapidly towards 250 degrees, slowed as it approached 250, then smoothly settled at 250. There was no overshoot or undershoot. It really seemed like magic, and of course I didn’t need to know any details of the math governing the PID controller.
Yes, please! I only understand the outline of the process which is that the generator frequency and phase must be made to match the power grid before the generator may be brought online. I would like to learn more about the process. In my electrical engineering courses, I learned about phase-locked-loops. In my physics work, I used synchronous detectors.
Those self tuning PID controls are cool stuff. We had to hire a tuner for our DCS, but then that was the three element drum level control.
Thanks for the nice historical context of the mechanical governor. The discussion of different power sources, steam versus gas turbine vs hydroelectric penstock etc. and the reasons not to modulate power too quickly or too slowly was nicely done. This could have been a very dull and dry description, but you made it interesting. Overall, a governor is a feedback device. In electronics, we are always worried about oscillation with feedback devices, and at least one portion of the loop must have a substantially slower response than the prime mover to avoid oscillation. One can achieve oscillation if you meet the Nyquist criterion, and I will stop there as it is too mathematical to be interesting in a short comment.
PS, you might enjoy this https://www.foleyengines.com/tech-tip-182-hoofpierce-governor-instructional-guide/#:~:text=ADJUSTING%20GOVERNOR%20*%20If%20the%20drop%20in,as%20necessary%20to%20achieve%20the%20desired%20performance.
Yikes! That is why the 100 KW 400 VDC generator we are designing and building for our two passenger 10 electric motor blown wing airplane uses electronic control of both the 143 HP turbonormalized engine and the active rectifier after the permanent magnet brushless generator. All those mechanical adjustments are too unreliable and too hard to modulate for flight.
Thank you, and yes oscillations are a problem, even in mechanical governors. It's one of the reasons you see droop in single units, the droop helps calm the mechanical action. As you take droop out the governor gets harder to tune. Most electronic governors are based on a PID loop, but it gets tricky because a governor swing can start swinging fuel pressure, and boost. Then you have all these other loops contributing to the problem. The PID triangle was there in mechanical and hydraulic governors too, but not as well defined.
For those readers not familiar with a PID controller, this a governor that uses the Position distance from a setpoint such as temperature, the Integral of the adjustment time from the deviation from the setpoint, and the Differential of rate of change returning to the setpoint. An example makes this more clear. I bought an industrial oven for a project that needed a precise 250° F to activate a single component epoxy to attach a metal disc to urethane. The urethane was rated to 250° F. The surplus oven did not work. I removed the old electronics and installed a tiny 2” by 2” learning PID temperature controller that had a digital display. The learning PID controller was industrial surplus and cost $100. I set it to 250° F. The first time was the learning run. The oven rapidly went to 251 degrees, then 249 degrees, then 250 degrees steady. The second time the oven was heated up the temperature went rapidly towards 250 degrees, slowed as it approached 250, then smoothly settled at 250. There was no overshoot or undershoot. It really seemed like magic, and of course I didn’t need to know any details of the math governing the PID controller.
I look forward to a discussion regarding synchronizing the phase of a generator to the interconnection prior to connection.
Gene! I completely missed this comment, SORRY! Did you want me to cover paralleling???
Yes, please! I only understand the outline of the process which is that the generator frequency and phase must be made to match the power grid before the generator may be brought online. I would like to learn more about the process. In my electrical engineering courses, I learned about phase-locked-loops. In my physics work, I used synchronous detectors.
Let me work on it, I probably need to do axtwo piece. 👍